Our model for the regulation of cholesterol biosynthesis is similar to general models for the action of steroid hormones. The model involves binding of an oxysterol metabolite to a specific binding protein. The sterol-protein complex formed then represses the synthesis of the regulatory enzyme in the pathway, 3-hydroxy-3-methlyglutary1-CoA (HMG-CoA) reductase. In support of this model, we have established the specificity of a cytosolic oxysterol binding protein for oxysterols that repress HMG-CoA reductase. Three different forms of the binding protein were detected: an unliganded form, Mr 236,000, which appears to be a trimer, a sterol-protein complex of Mr 169,000 which is a dimer, and a monmeric form of the complex, Mr 97,000. At acid pH (below 6.0), the protein undergoes a conformational change which greatly increases rates of sterol binding and dissociation and which allows the protein to bind nonspecifically to DNA. The binding protein has been partially purified (about 200-fold). We have identified two oxysterols, 24(S),25-epoxycholesterol and 25-hydroxycholesterol, in cultured fibroblasts and have demonstrated that both are produced in concentrations within the range that is required to regulate HMG-CoA reductase and sterol synthesis. 24(S),25-Epoxycholesterol arises via a branch in the pathway beginning with the formation of squalene 2,3;22,23-dioxide. We do not yet know the reactions that give rise to 25-hydroxycholesterol. Preliminary evidence indicates that the two oxysterols can be metabolized to more polar sterols which do not bind to the bnding protein. (D)